Liu et al. J Transl Med (2020) 18:130 https://doi.org/10.1186/s12967-020-02282-3 Journal of Translational Medicine

RESEARCH Open Access Characteristics of the urinary microbiome in stone patients with Fengping Liu1,2†, Nan Zhang2†, Peng Jiang2†, Qixiao Zhai3†, Chen Li2, Deshui Yu2, Yan Wu2, Yuwei Zhang2, Longxian Lv4*, Xinyu Xu2* and Ninghan Feng2*

Abstract Background: Kidney stone (KSD) is more common in individuals with hypertension (HTN) than in individuals with normotension (NTN). Urinary dysbiosis is associated with urinary tract disease and systemic . However, the role of the urinary microbiome in KSD complicated with HTN remains unclear. Methods: This study investigated the relationship between the urinary microbiome and blood pressure (BP) in patients with KSD co-occurring with HTN (KSD-HTN) and healthy controls (HC) by conducting 16S rRNA gene sequencing of in samples. The urine samples were collected (after bladder disinfection) from 50 patients with unilateral kidney stones and NTN (n 12), prehypertension (pHTN; n 11), or HTN (n 27), along with 12 HCs. = = = Results: Principal coordinates analysis showed that there were signifcant diferences in the urinary microbiomes not only between KSD patients and HCs but also between KSD-pHTN or KSD-HTN patients and KSD-NTN patients. Gardnerella dominated in HCs, Staphylococcus dominated in KSD-NTN patients and Sphingomonas dominated in both KSD-pHTN and KSD-HTN patients. The abundance of several genera including Acidovorax, Gardnerella and Lactobacil- lus was correlated with BP. Adherens junction and nitrogen and nucleotide pathways, among others, were associated with changes in BP. Conclusions: The fndings suggest that patients with KSD complicated with HTN have a unique urinary microbiome profle and that changes in the microbiome may refect disease progression and may be useful to monitor response to treatments. Keywords: Kidney pelvis, , Microbiome, Hypertension, Prehypertension, Urinary bacteria

Background Kidney stone disease (KSD) is common, with a preva- lence of up to 14.8%, which is increasing over time, and a recurrence rate of up to 50% within the frst 5 years of *Correspondence: [email protected]; [email protected]; the initial episode [1]. KSD disproportionately afects [email protected] †Fengping Liu, Nan Zhang, Peng Jiang and Qixiao Zhai contributed patients with hypertension (HTN) compared to individu- equally to this work als with normal blood pressure (BP), i.e. normotension 2 Department of , Afliated Wuxi No. 2 Hospital, Nanjing Medical (NTN) [2]. Both KSD and HTN impair kidney function; University, Wuxi, Jiangsu, China 4 State Key Laboratory for Diagnosis and Treatment of Infectious Diseases, this can lead to chronic , which negatively National Clinical Research Center for Infectious Diseases, Collaborative afects quality of life and can be fatal [3, 4]. Innovation Center for Diagnosis and Treatment of Infectious Diseases, Clarifying the shared pathophysiologic basis of KSD The First Afliated Hospital, College of Medicine, Zhejiang University, Hangzhou, Zhejiang, China and HTN could lead to more efective treatments for Full list of author information is available at the end of the article patients. In KSD patients, HTN was previously found to

© The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativeco​ ​ mmons.org/licen​ ses/by/4.0/​ . The Creative Commons Public Domain Dedication waiver (http://creativeco​ mmons​ .org/publi​ cdoma​ in/​ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Liu et al. J Transl Med (2020) 18:130 Page 2 of 13

be associated with signifcantly increased urine calcium this hypothesis by comparing the urinary microbiome [5], which may result in low blood calcium. Adequate cal- profles (based on bacterial 16S rRNA gene sequenc- cium levels may regulate BP by modifying intracellular ing of urine samples) of healthy controls (HCs) and KSD calcium in vascular smooth muscle cells and by varying patients with NTN, prehypertension (pHTN) and HTN. the blood volume via the renin–angiotensin–aldoster- Te results demonstrate that the co-occurrence of HTN one system. Low blood calcium can increase the activ- alters the urinary microbiome composition of KSD ity of the parathyroid gland, and patients, which has important implications for disease increases intracellular calcium in vascular smooth mus- diagnosis and management. cles, resulting in vasoconstriction. Low blood calcium also increases the synthesis of calcitriol in a direct man- Methods ner or via parathyroid hormone, and calcitriol increases Subject recruitment intracellular calcium in vascular smooth muscle cells. Kidney stone disease patients who were undergoing ure- Low blood calcium may stimulate renin release and teroscopic were recruited and classifed into consequently increase the levels of angiotensin II and the following three groups according to BP: KSD-NTN, aldosterone [6], which are responsible for regulating BP. KSD-pHTN and KSD-HTN. Te presence of kidney Recent studies revealed that angiotensin II-induced HTN stones was confrmed by abdominal X-ray, ultrasonog- is associated with gut microbial composition and metab- raphy and computed tomography (CT), and calcium olites [7, 8]. stones were identifed by CT scans. Normal BP was Research has shown that human urine harbors its own defned as a systolic blood pressure (SBP) ≤ 120 mmHg microbial community, which has challenged the long- or a diastolic blood pressure (DBP) ≤ 80 mmHg; pHTN held notion that urine is sterile in the absence of infection was defned as an SBP of 120–139 mmHg or a DBP of [9]. Just as the gut, oral cavity and vaginal microbiomes 80–89 mmHg without the use of antihypertensive medi- contribute to human health, the urinary microbiome is cation; and HTN was defned as an SBP ≥ 140 mmHg or critical for the maintenance of physiologic homeostasis, a DBP ≥ 90 mmHg and/or use of antihypertensive medi- as demonstrated by the urinary dysbiosis observed in cation [20]. In addition, a cohort of subjects with neither prostate cancer [10], bacterial vaginosis [11] and neuro- stones nor HTN was recruited as the HC group. Exclu- pathic bladder [12]. sion criteria for patients and HCs were as follows: men- Te urinary metabolite profle of KSD patients with struation, pregnancy, cancer, heart failure, renal failure, HTN difers from that of KSD patients with NTN. For peripheral artery disease, urinary tract disease, urinary example, it was reported that , , titrat- tract deformity, known (UTI) able acid and ammonium were increased in the urine based on clinical assessment, urinary catheterization of patients with calcium stones plus HTN compared to within the previous 4 weeks and treatment with antibiot- calcium stones plus NTN, whereas urinary pH and cit- ics within the previous 4 weeks. rate were decreased [13–15]. Tese changes can afect the urinary environment and, consequently, the relative Urine sample collection and DNA extraction abundance of bacteria in the urinary microbiome. Along Urine sample collection procedures for the KSD patients with being present in bladder urine, microorganisms were as follows: bladder urine was aspirated using a cys- populate urine in the upper urinary tract. toscope (Richard Wolf, Knittlingen, Germany) and the Recent studies have indicated the plausible involve- bladder was then disinfected three times with iodophor; ment of human microbiomes (in various body sites) in 2 mL of the last iodophor fush was used for expanded the regulation of BP. For example, in HTN patients, both quantitative urine culture (EQUC) [21] and patients with Yang et al. and Li et al. found that BP was associated with a positive EQUC (> 10 colony-forming units/mL of urine) dramatically decreased gut microbial richness and diver- were excluded [21]. After the last iodophor fush, nor- sity and bacterial composition [16, 17]. In another study mal saline was injected into the bladder and immediately on the oral microbiome, Ko et al. [18] demonstrated that aspirated to remove any iodophor remaining in obstructive sleep apnea/hypopnea syndrome (OSAHS) the bladder. Tis step was repeated three times and 3 mL patients with HTN had a diferent bacterial profle from urine was then aspirated from the kidney pelvis using a those without HTN. ureteroscope. As the procedure for the collection of kid- Given the interactions between the human microbiome ney pelvis urine samples was too invasive to obtain con- and the metabolome [19] and between the gut and oral sent from HCs and ethics approval from the hospital microbiomes and BP [16–18], we speculated that the uri- ethics committee, bladder urine samples were collected nary microbiome composition is associated with difer- from the HCs by transurethral catheterization instead. ences in BP and metabolism in KSD patients. We tested Specifcally, the catheter was inserted to the bladder and Liu et al. J Transl Med (2020) 18:130 Page 3 of 13

withdrawn after 5 mL was collected. Urine samples were diferences in the microbial communities between sam- immediately stored at − 80 °C until further processing. ples. Using the vegan package in R, we applied the per- For DNA extraction, 1 mL urine was centrifuged at mutational multivariate analysis of variance method to 20,000×g for 30 min and the pellet was resuspended in the Bray–Curtis distance data using 1000 permutations 150 μL lysis bufer (BGI Group, Shenzhen, China). Sera- to analyze diferences in OTUs between the four groups. Mag SpeedBeads Carboxylate-Modifed Magnetic Parti- Statistical signifcance was defned as p < 0.05. Based cles (GE Healthcare UK, Little Chalfont, UK) were used on the OTU abundances, a Venn diagram was used to to extract the DNA, as described in our previous study display the numbers of microbial OTUs shared by the [22]. Te DNA concentration was measured using a various groups. Functional analysis of microbiomes asso- Qubit Fluorometer (Termo Fisher Scientifc, Waltham, ciated with the three BP categories was carried out using MA, USA). Tereafter, PCR amplifcation (involving 35 PICRUSt [25]. cycles) was conducted using the universal primers 341F and 806R, which target the V3–V4 region of the 16S Statistical analysis rRNA gene. Amplicons were analyzed by gel electropho- Te mean relative abundances of genera and phyla in resis and purifed using a QIAquick gel extraction kit each group were used to describe the urinary microbi- (Qiagen, Hilden, Germany). Te purifcation product was ome compositions. One-way analysis of variance was diluted to 10 ng/μL, and 5 μL of each sample was then used to compare the quantitative clinical variables among pooled for PE300 sequencing on a HiSeq 2500 system the four groups. Wilcoxon rank-sum test was used to (Illumina, San Diego, CA, USA). compare alpha diversity indices and bacterial abundances A negative control consisting of normal saline was used among the four groups. Statistical signifcance was to assess the contribution of contaminating DNA from defned as p < 0.05 [26]. Te tables and fgures display the the reagents, and a negative control without template mean and standard deviation (SD) of each variable. In the DNA was included in the sequencing protocol. Nine fgures, the horizontal bar represents the mean and the samples per batch were sequenced in duplicate to con- error bar represents ± SD. Pearson correlation analysis frm the reproducibility of the results. was used to assess the correlations between the relative abundances of bacterial genera and both SBP and DBP Bioinformatic analysis in all participants. Data were analyzed using SPSS v.24.0 Paired-end reads were assigned to samples based on their (SPSS Inc., Chicago, IL, USA). unique barcode, which was then removed along with the primer sequence. Te paired-end reads were merged Results using FLASH [23]. Quality fltering of raw tags was per- Demographic and clinical variables formed under specifc fltering conditions to obtain As shown in Table 1, urine samples were collected from high-quality clean tags using fqtrim v.0.94. Chimeric HCs (n = 12) and KSD-NTN (n = 12), KSD-pHTN sequences were fltered out and sequences with ≥ 97% (n = 11) and KSD-HTN (n = 27) patients. As expected, similarity were assigned to the same operational taxo- the SBP and DBP in the KSD-pHTN and KSD-HTN nomic unit (OTU) using Vsearch v.2.3.4. A representative groups were signifcantly higher than those in the HC sequence was selected for each OTU and taxonomic data and KSD-NTN groups (p < 0.05). Te number of males were assigned to each representative sequence using the and difered signifcantly among the Ribosomal Database Project classifer with a confdence four groups (p < 0.05), whereas drinking and smoking his- value of 0.8 as the cutof. Samples with < 30,000 clean tory, UTI, comorbidities, creatinine, blood tags were removed. nitrogen, estimated glomerular fltration rate and urinary Contaminant sequences (based on the negative con- pH were not signifcantly diferent among the four groups trols) were removed using Decontam v.1.2.1 with p < 0.10 (p > 0.05), and duration of stones were similar among the as the threshold [24]. Bray–Curtis dissimilarity was used three KSD groups (p > 0.05). to quantify diferences in OTUs to confrm the reproduc- ibility of the duplicate sequenced samples. 16S gene sequence‑based characterization of patient OTU abundance data were normalized using a stand- groups ard sequence number corresponding to the sample with Two samples in the KSD-HTN group were negative for the smallest number of sequences. Alpha diversity was the 16S rRNA gene whereas all samples in the other used to analyze the complexity of species diversity in each groups were positive. Te samples yielded a total of 2346 sample, which involved using QIIME v.1.8.0 to calculate OTUs (509, 1126, 1009 and 1738 in the HC, KSD-NTN, Chao1, Observed species, Shannon index and Simpson’s KSD-pHTN and KSD-HTN groups, respectively). index. Beta diversity analysis was performed to evaluate Liu et al. J Transl Med (2020) 18:130 Page 4 of 13

Table 1 Demographic and clinical characteristics of the study population Characteristic HC (n 12) Value for cohort ­(na)b or statistic p value­ c = KSD-NTN (n 12) KSD-pHTN (n 11) KSD-HTN (n 27) = = = Male [no. (%)] 9 (75.00) 6 (50.00) 4 (36.00) 23 (85.10) 0.013 Married status [no. (%)] 12 (100.00) 12 (100.00) 11 (100.00) 27 (100.00) 1.000 Age 58.91 18.97 47.33 14.95 54.09 13.03 54.74 12.36 0.270 ± ± ± ± Body mass index (kg/m2) 25.08 3.33 22.67 2.13 24.38 1.69 25.82 2.80 0.011 ± ± ± ± History of drinking 0 (0.00) 1 (8.33) 2 (18.18) 1 (3.70) 0.291 History of smoking 0 (0.00) 1 (8.33) 1 (9.09) 4 (14.81) 0.546 Urinary tract infection in previous 1 month 0 (0.00) 1 (8.33) 0 (0.00) 0 (0.00) 0.340 Duration of stones (years) 0.375 < 0.5 year [no. (%)] NA 11 (91.67) 10 (90.91) 25 (92.59) 0.5 to 1 year [no. (%)] NA 1 (8.33) 1 (9.09) 0 (0.00) > 1 year [no. (%)] NA 0 (0.00) 0 (0.00) 2 (7.41) Duration of hypertension (years) NA NA NA 4.48 4.59 NA ± Systolic blood pressure (mmHg) 121.83 5.52 113.58 8.35 131.45 7.46 151.11 9.68 0.000 ± ± ± ± Diastolic blood pressure (mmHg) 74.92 7.79 70.33 4.64 80.64 7.06 92.00 9.74 0.000 ± ± ± ± Comorbidities Type 2 diabetic mellitus [no. (%)] 0 1 (8.33) 4 (36.36) 5 (18.52) 0.055 Coronary heart disease [no. (%)] 0 0 (0.00) 0 (0.00) 2 (7.41) 0.332 Serum creatinine (μmol/L) 68.90 10.77 72.89 21.90 73.78 28.80 93.56 40.36 0.069 ± ± ± ± Blood urea nitrogen (mmol/L) 5.75 1.61 5.75 1.14 5.06 1.43 6.20 2.25 0.387 ± ± ± ± Estimated glomerular fltration rate (mL/min/1.73 m2) 97.45 13.96 95.58 20.95 89.91 21.80 82.89 26.19 0.209 ± ± ± ± Urinary pH 6.34 0.65 6.38 0.88 5.77 0.82 6.39 1.12 0.314 ± ± ± ± HC healthy controls, HTN hypertension, KSD kidney stone disease, NA not applicable, NTN normotension, pHTN pre-hypertension a n, no. of subjects b Mean SD or no. (%) ± c Fisher’s exact test was used for categorical variables and one-way analysis of variance was used to compare continuous variables

We calculated the Chao1 and Observed species indices 1875 OTUs in the HC and KSD-HTN samples, of which to evaluate the richness of the urinary microbiome in the 372 (19.84%) were shared by the two groups; there were four groups and the Shannon index and Simpson’s index 1590 OTUs in the KSD-NTN and KSD-pHTN samples, to assess diversity (Fig. 1a). HC samples had signifcantly of which 545 (34.28%) were shared by the two groups; decreased indices of bacterial richness and diversity com- there were 2056 OTUs in the KSD-NTN and KSD-HTN pared to those in the other three groups (p < 0.05). In samples, of which 808 (39.30%) were shared by the two addition, the KSD-HTN samples tended to have higher groups; and there were 1976 OTUs in the KSD-pHTN bacterial richness and diversity than the KSD-NTN and and KSD-HTN samples, of which 771 (39.02%) were KSD-pHTN samples, but the diferences were non-signif- shared by the two groups. icant (p > 0.05). Principal coordinates analysis (PCoA) revealed sig- Diferential abundances of bacterial phyla among groups nifcant diferences in bacterial composition between the Te composition of bacterial phyla difered between HC and KSD-NTN groups (p = 0.002), the HC and KSD- the four groups. Specifcally, 14, 16, 17 and 23 phyla pHTN groups (p = 0.002), the HC and KSD-HTN groups were detected in the HC, KSD-NTN, KSD-pHTN and (p = 0.002), the KSD-NTN and KSD-pHTN groups KSD-HTN groups, respectively. Te dominant bacte- (p = 0.024) and the KSD-NTN and KSD-HTN groups rial phylum in the HC group was Bacteroidetes (29.20%), (p = 0.024) (Fig. 1b). followed by Proteobacteria (28.49%) and Actinobacteria As shown in the Venn diagram in Fig. 1c, there were (23.07%). In the KSD-NTN group, the major phylum was 1331 OTUs in the HC and KSD-NTN samples, of which Proteobacteria (46.07%), followed by Firmicutes (30.37%) 304 (22.84%) were shared by the two groups; there were and Actinobacteria (10.70%). In the KSD-pHTN group, 1228 OTUs in the HC and KSD-pHTN samples, of which the major phyla were Proteobacteria (33.38%), Firmi- 290 (23.62%) were shared by the two groups; there were cutes (27.10%) and Bacteroidetes (14.84%). Lastly, in the Liu et al. J Transl Med (2020) 18:130 Page 5 of 13

a

C N N N C N N C N N C N H T T H T TN H T H H N HT HT NT -NT p -H p H - - D - C -p C D -pHTNC-HTN H HC-H SD-NTNC H S H KS HC KSD- HC- K H K HC

b PCoA (p = 0.001) c Venn

KSD-NTN KSD-pHTN

HC KSD-HTN

Fig. 1 Bacterial community structure in HC and kidney stone disease (KSD) patients with normotension, prehypertension and hypertension (KSD-NTN, KSD-pHTN and KSD-HTN, respectively). a Bacterial richness and diversity across groups. Comparison of urinary microbiome alpha richness and diversity (Chao1, Observed species, Shannon index and Simpson’s index) between HC and KSD-NTN, between HC and KSD-pHTN, between HC and KSD-HTN, between KSD-NTN and KSD-pHTN, between KSD-NTN and KSD-HTN and between KSD-pHTN and KSD-HTN, using Wilcoxon rank-sum test. Horizontal bar represents mean and error bar represents SD. Bacterial richness and diversity were signifcantly greater ± in the HC than those in the three KSD groups (p < 0.05), and were slightly greater in the KSD-HTN group than those in the other two KSD groups (p > 0.05). b Principal coordinates analysis (PCoA) revealed the clustering of bacterial taxa in the three groups based on Bray–Curtis distance, with each point corresponding to a subject and colored according to the sample type. Permutational multivariate analysis of variance showed that the separation of bacterial communities in the four groups was signifcant (p 0.001); the separation was also signifcant in HC vs KSD-NTN (p 0.002), = = HC vs KSD-pHTN (p 0.002), HC vs KSD-HTN (p 0.002), KSD-NTN vs KSD-pHTN (p 0.024) and KSD-NTN vs KSD-HTN (p 0.024), but not in = = = = KSD-pHTN vs KSD-HTN (p 0.111). c Venn diagram showing a dissimilar number of operational taxonomic units shared by HC and KSD-NTN, by HC = and KSD-pHTN, by HC and KSD-HTN, by KSD-NTN and KSD-pHTN, by KSD-NTN and KSD-HTN and by KSD-pHTN and KSD-HTN. KSD kidney stone disease, HC healthy controls, HTN hypertension, NTN normotension, pHTN pre-hypertension

KSD-HTN group, Proteobacteria (31.79%), Firmicutes Diferential abundances of bacterial genera among groups (30.47%) and Actinobacteria (20.94%) were predomi- We detected 238, 329, 326 and 474 genera in the HC, nant (Fig. 2a). Moreover, Acidobacteria and Deinococ- KSD-NTN, KSD-pHTN and KSD-HTN groups, respec- cus–thermus were enriched in the KSD-pHTN group tively. Te dominant genera in the HC group were compared to the KSD-NTN and HC groups, and Deino- Gardnerella (17.66%), Pontibacter (8.50%), Sphingo- coccus–thermus was enriched in the KSD-HTN group monas (6.87%), Prevotella (2.41%) and Propionibacte- compared to the HC group. Bacteroidetes was depleted rium (1.57%). In the KSD-NTN group, the dominant in both the KSD-NTN and KSD-HTN groups compared genera were Staphylococcus (10.09%), Sphingomonas to the HC group. Fusobacteria was enriched in the KSD- (8.34%), Delftia (5.89%), Acinetobacter (5.69%) and pHTN group compared to the HC and KSD-HTN groups Prevotella (2.68%). In the KSD-pHTN group, Sphingo- (Additional fle 1: Figure S1). monas (7.97%) was the most highly represented genus, followed by Pontibacter (7.08%), Acinetobacter (4.64%), Liu et al. J Transl Med (2020) 18:130 Page 6 of 13

Phylum Genus

100 100 abOthers Prevotella Corynebacterium %) %) e( e( Others Delftia Deinococcus-Thermus Propionibacterium Bacteroidetes undanc undanc 50 50 Staphylococcus ab ab Actinobacteria Bifidobacterium Firmicutes Gardnerella Proteobacteria Acinetobacter Relative Relative Pontibacter Sphingomonas 0 0

C N C H TN TN T TN TN TN H H -N -H -N H -H D -p D D -p D S D S S D S K S K K S K K K Fig. 2 Distribution of bacterial phyla and genera in HC and kidney stone disease (KSD) patients with normotension, prehypertension and hypertension (KSD-NTN, KSD-pHTN and KSD-HTN, respectively). a, b Relative abundances of the major bacterial phyla (a) and genera (b) were determined based on 16S rDNA gene sequencing. “Other” refers to the combined relative abundance for all taxa not including the top fve most abundant phyla and top ten most abundant genera. KSD kidney stone disease, HC healthy controls, HTN hypertension, NTN normotension, pHTN pre-hypertension

Propionibacterium (4.53%) and Corynebacterium and Stenotrophomonas had the lowest abundance in the (4.15%). Sphingomonas was also the most common genus KSD-pHTN group (Fig. 3 and Additional fle 2: Table S1). in the KSD-HTN group (8.63%), followed by Bifdobac- However, several bacteria gradually increased with the terium (7.04%), Acinetobacter (3.99%), Propionibacterium KSD patients’ BP, including Pseudomonas (0.72% in the (3.68%) and Delftia (2.87%). Although Staphylococcus KSD-NTN group, 0.97% in the KSD-pHTN group and was the most abundant genus in the KSD-NTN group 3.03% in the KSD-HTN group). (10.09%), it accounted for only 2.15% and 1.52% of genera In total, 20 bacterial genera exhibited signifcant difer- in the KSD-pHTN and KSD-HTN groups, respectively ences between the KSD-NTN and KSD-pHTN groups. (Fig. 2b). For example, Achromobacter, Clostridium IV and Delftia We also compared the relative abundances of each were enriched while Arthrobacter, Enhydrobacter and genus among the four groups. Compared to the HC Pontibacter were depleted in the KSD-NTN group com- group, there were 51, 30 and 37 genera with signif- pared to the KSD-pHTN group (Fig. 3). cantly diferent relative abundances in the KSD-NTN, Tere were 31 bacterial genera that exhibited signif- KSD-pHTN and KSD-HTN groups, respectively. Genera cant diferences between the KSD-NTN and KSD-HTN exhibiting signifcant diferences among the three KSD groups. For example, Acidovorax, Bacteroides and Delf- groups are shown in Fig. 3 and genera exhibiting signif- tia were enriched while Arthrobacter, Cloacibacterium cant diferences between the HC and KSD groups are and Lactobacillus were depleted in the KSD-NTN group shown in Additional fle 2: Table S1. For example, Coma- compared to the KSD-HTN group (Fig. 3). monas was enriched in the KSD subjects compared to the Tere were 16 genera that exhibited signifcant difer- HC group, Enterococcus was enriched in the KSD-NTN ences between the KSD-pHTN and KSD-HTN groups. and KSD-HTN groups compared to the HC group and For example, Anaerovorax, Oerskovia and Streptomyces Bifdobacterium and Lactobacillus were enriched in the were depleted while Aureimonas, Pontibacter and Veil- KSD-pHTN group compared to the HC group (p < 0.05). lonella were enriched in the KSD-pHTN group com- Notably, several genera showed a parabolic trend pared to the KSD-HTN group (Fig. 3). across the three stages of BP. For example, both Bifdo- bacterium and Lactobacillus had the highest abundance Bacterial community composition is correlated with BP in the KSD-HTN group, the lowest abundance in the To explore the associations of bacteria with SBP and KSD-pHTN group and intermediate abundance in the DBP, we conducted correlation analyses using all of the KSD-NTN group. In contrast, Ochrobactrum, Rhizobium participants. Some bacterial genera were correlated with Liu et al. J Transl Med (2020) 18:130 Page 7 of 13

Achromobacter Acidovorax Amaricoccus Anaerovorax Arthrobacter Asticcacaulis Aureimonas

1.5 * 1.0 * 0.20 * 0.8 * 4 * 0.4 * 40 * *** * * %) %) %) %) 0.8 %) %) %) e( e( e( e( 0.15 e( 0.6 e( 3 0.3 e( 30 1.0 danc danc danc danc 0.6 danc danc danc n n 0.10 n 0.4 n 2 0.2 n 20 ab un ab un abu abu 0.4 abu abu abu tive tive tive 0.5 tive tive tive tive la la la la 0.05 la 0.2 la 1 0.1 la 10 0.2 Re Re Re Re Re Re Re

0.0 0.0 0.00 0.0 0 0.0 0

N N N N N N N N N TN TN TN TN TN TN T T T TN T T T T TN TN TN TN TN -N H -H -N H -H -NT H -HT -N H -H -N H -H -N H -H -N H -H D -p D D -p D D -p D D -p D D -p D D -p D D -p D S D S S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K S K K K K K K K K

Bacteroides Bilophila Blastococcus Bradyrhizobium Brevibacterium Butyricicoccus Butyrivibrio

10 * 0.3 * 0.3 * 2.0 * 0.5 * 0.5 * 0.15 * * ) %) %) %) %) %) %)

8 (% 0.4 0.4 e( e( e( 1.5 e( e( e( 0.2 0.2 0.10 danc danc danc dance 6 danc 0.3 danc 0.3 danc un 1.0 un un un ab un ab un ab un ab 4 ab 0.2 ab 0.2 ab tive tive tive tive 0.1 0.1 tive tive tive 0.05 la la la la 0.5 la la la 2 Re Re Re Re Re 0.1 Re 0.1 Re

0 0.0 0.0 0.0 0.0 0.0 0.00

N N N N N N N N N N TN T T TN T TN TN TN TN T TN T TN TN T TN T TN -N H -H -N H -HT -N H -H -NT H -HT -N H -H -N H -H -N H -H D -p D D -p D D -p D D -p D D -p D D -p D D -p D S D S S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K S K K K K K K K K

Cardiobacterium Cellvibrio Cloacibacterium Clostridium Clostridium IV Comamonas Delftia

0.15 * 0.5 * 8 * 0.4 * 0.6 ** 1.5 * 30 * * * %) %) %) %) %) %) %) 0.4 e( e( e( e( e( e( e( 6 0.3 0.10 0.4 1.0 20 danc danc danc danc danc danc danc 0.3 4 0.2 ab un ab un ab un ab un ab un ab un ab un 0.2 tive tive tive tive tive tive tive 0.05 0.2 0.5 10 la la la la la la la 2 0.1 Re Re Re Re Re Re Re 0.1

0.00 0.0 0 0.0 0.0 0.0 0

N N N N N N N N N N N N N N T T T T TN T T T T TN T T TN TN TN TN T TN -N -H -N H -HT -N H -H -N H -H -N H -HT -N H -H -N -HT D -pH D D -p D D -p D D -p D D -p D D -p D D -pH D S D S S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K S K K K K K K K K

Edaphobacter Enhydrobacter Flaviflexus Lachnospiracea incertae sedis Lactobacillus Methylophilus Neorhizobium

0.10 * 3 * 0.5 * 6 * 5 * 0.6 * 0.20 ** * * ) ) ) ) ) ) ) (% (% (% 0.08 (% 0.4 (% 4 (% (% 0.15 2 4 0.4 0.06 0.3 3 ndance ndance ndance ndance ndance ndance ndance 0.10 abu abu abu 0.04 abu 0.2 abu 2 abu abu tive tive tive tive 1 tive 2 tive 0.2 tive la la la la la la la 0.05 Re Re Re 0.02 Re 0.1 Re 1 Re Re

0.00 0 0.0 0 0 0.0 0.00

N N N N N N N N N N N N N TN TN T TN TN TN T T T TN T T T TN T TN -N H -HT -NT H -H -NT H -H -N H -HT -N H -H -N H -HT -N H -H D -p D D -p D D -p D D -p D D -p D D -p D D -p D S D S S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K S K K K K K K K K

Ochrobactrum Oerskovia Paracoccus Parasutterella Pelomonas Phascolarctobacterium Planococcus

3 ** 3 * 3.5 * 1.5 * 4 * 0.4 ** 12 * * * * 3.0 %) %) %) %) %) %) %) 10 e( e( e( e( e( e( 2.5 3 0.3 e( 2 2 1.0 8 ndanc ndanc ndanc ndanc ndanc ndanc 2.0 ndanc 2 0.2 6 abu abu abu abu abu abu 1.5 abu tive tive tive tive tive tive 1 1 0.5 tive 4 la la la la la la 1.0 1 0.1 la Re Re Re 0.5 Re Re Re Re 2 0 0 0.0 0.0 0 0.0 0

N N N N N N N N N N T TN TN TN T TN TN T TN TN T TN TN T T TN T TN -NT -H -NT H -H -N H -HT -N H -H -N -H -N -H -N H -H D -pH D D -p D D -p D D -p D D -pH D D -pH D D -p D S D S S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K S K K K K K K K K

Pontibacter Propionibacterium Pseudochrobactrum Ramlibacter Rhizobium Rhodanobacter Rhodococcus

20 * 10 * 10 ** 0.25 * 10 ** 3 * 0.8 * * * ** %) %) %) %) %) %) %) 8 8 0.20 8 e( e( e( e( e( e( 15 e( 0.6 2 danc danc danc danc danc danc danc 6 6 0.15 6 10 0.4 ab un ab un ab un ab un ab un ab un ab un 4 4 0.10 4 tive tive tive tive tive tive tive 1 la la la la la la 5 la 0.2 Re Re Re Re Re Re Re 2 2 0.05 2

0 0 0 0.00 0 0 0.0

N N N N N N N N N N N N TN T T T TN T T TN TN TN TN T TN T T T TN TN -N H -H -N H -H -NT H -HT -N H -H -N H -H -N H -H -NT H -H D -p D D -p D D -p D D -p D D -p D D -p D D -p D S D S S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K S K K K K K K K K

Selenomonas Sphingobium Stenotrophomonas Streptomyces Succiniclasticum Syntrophococcus Vampirovibrio Veillonella

0.15 * 1.5 * 4 ** 0.5 * 0.6 * 0.20 * 1.0 * 1.0 * * %) %) %) %) %) %) 0.4 %) %) 0.8 0.8 e( e( e( e( e( e( 3 e( 0.15 e( 0.10 1.0 0.4

danc 0.6 danc danc danc danc danc 0.3 danc danc 0.6 2 0.10 ab un ab un ab un ab un ab un ab un 0.2 ab un ab un 0.4 0.4 tive tive tive tive tive tive 0.05 0.5 tive 0.2 tive la la la la la la 1 la 0.05 la

Re 0.2 Re Re Re Re Re 0.1 Re Re 0.2

0.00 0.0 0 0.0 0.0 0.00 0.0 0.0 N N N N N N N N N N N N N N N N N T TN TN TN T T TN T T TN T TN T T T T TN H -N H -HT -NT H -H -N -H -NT H -HT -N H -H -NT H -HT -NT H -H -N pH - D -p D D -p D D -pH D D -p D D -p D D -p D D -p D D - D S D S S D S S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K S K K S K K K K K K K K K Fig. 3 Bacterial abundance showing signifcant diference between three stages of blood pressure in KSD patients. Wilcoxon rank-sum test was used to compare the diference of abundance between two groups. *, **, *** means p < 0.05, p < 0.01, p < 0.001, respectively. HTN hypertension, KSD kidney stone disease, NTN normotension, pHTN pre-hypertension Liu et al. J Transl Med (2020) 18:130 Page 8 of 13

in the gene content of the urinary microbes not only a b between the HC group and the three KSD groups, but SBP r DBP r also between the KSD-NTN and KSD-pHTN groups, the KSD-NTN and KSD-HTN groups and the KSD-pHTN and KSD-HTN groups. Specifcally, compared to the KSD-pHTN group, lipid biosynthesis and nucle- otide metabolism pathways were underrepresented and mRNA surveillance, nitrogen metabolism, other trans- porters and pancreatic secretion pathways were enriched in the KSD-NTN group (p < 0.05; Fig. 5 and Additional fle 3: Table S2). Compared to the KSD-HTN group, adherens junction, nitrogen metabolism and pancreatic secretion pathways were enriched and metabolism of cofactors and vitamins and nucleotide metabolism path- ways were depleted in the KSD-NTN group (p < 0.05; Fig. 5 and Additional fle 3: Table S2). Compared to the KSD-HTN group, ATP-binding cassette transporter and mRNA surveillance pathways were underrepresented in the KSD-pHTN group (p < 0.05; Fig. 5 and Additional fle 3: Table S2).

Discussion Tis study investigated whether the urinary microbiome of KSD patients difers according to BP. Signifcant difer- ences in bacterial community composition were observed between KSD patients with normal and abnormal BP. We also found that HTN-associated microbial communities Fig. 4 Pearson correlation analysis of bacterial genera and blood exhibited alterations in metabolic pathways. Tese dif- pressure (BP). a, b Bacterial genera most closely correlated with SBP ferences indicate that there are additional risk factors for (a) and DBP (b) in all subjects in the present study are shown. Positive and negative values of r indicate positive (red) and negative (blue) HTN co-occurrence with KSD and some of these difer- correlations, respectively, between the relative abundance of a genus ences are potential targets for therapeutic interventions. and SBP or DBP. Only signifcant correlations (p < 0.05) are shown. DBP Based on the estimators of bacterial richness and diver- diastolic blood pressure, SBP systolic blood pressure sity, there were more detectable bacterial phyla and gen- era in KSD patients’ urine samples than the bladder urine samples from HCs. In addition, the richness and diversity estimators in the KSD-HTN group were non-signifcantly both SBP and DBP, such as Acidovorax, Anaerovorax and slightly higher than those in the other two KSD groups. Cellulosilyticum (Fig. 4a, b). Interestingly, there were It was previously reported that the gut bacterial richness some bacterial genera that correlated diferently with and diversity in individuals with pHTN or HTN with- SBP vs DBP. For example, Corynebacterium, Enterococ- out KSD were dramatically lower than those in HCs, but cus and Lactobacillus were positively correlated with SBP the alterations in their urine samples, which would help (Fig. 4a) but not with DBP. In contrast, Bilophila, Gard- to understand the role of KSD in HTN, have not previ- nerella and Rhizobium were negatively correlated with ously been reported [16]. Although low bacterial diver- SBP (Fig. 4a) but not with DBP. Acholeplasma, Anaero- sity was associated with lack of kidney stones and low plasma and Geobacillus were positively correlated with BP in the present study, it is not rational to conclude that DBP but not with SBP (Fig. 4b). In contrast, Butyrivibrio, low bacterial diversity indicates a healthy status, as previ- Marinobacter and Succiniclasticum were negatively cor- ous studies on the human urinary microbiome have not related with DBP (Fig. 4b) but not with SBP. reached a consistent conclusion. For example, Wu et al. Metabolic biosynthesis pathways associated found that women with had lower with the urinary microbiome and BP bacterial diversity and richness than HCs [27], whereas Tomas-White et al. reported that increased microbiome To assess the gene content of the urinary microbes, pre- diversity was associated with urgent dictions were made using PICRUSt based on a micro- symptoms [28] and Wu et al. found that male patients bial genome database. Tere were signifcant diferences Liu et al. J Transl Med (2020) 18:130 Page 9 of 13

Adherens junction ATP-binding cassette transporter Hypertrophic cardiomyopathy Leukocyte transendothelial migration Lipid biosynthesis proteins

800 * 15000 * 800 * 800 * 150000 *

600 600 600 10000 100000

400 400 400

5000 50000 200 200 200

0 0 0 0 0

N N N N N T TN TN TN T TN TN -N -HTN -NTN -HT -NT -H -NTN H -HTN -NT H -HTN Relativeabundance(%) Relativeabundance(%) D -pH D Relativeabundance(%) D -pH D Relativeabundance(%) D -pH D Relativeabundance(%) D -p D D -p D S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K K K K K

Metabolism of cofactors and vitamins mRNA surveillance Nitrogen metabolism Nucleotide metabolism Other transporters Pancreatic secretion ) ) * 4000 ** * * * * 2000 * 15000 * 2500 * 1500 * 2500 * 2000 2000 1500 3000 10000 1000 1500 1500 1000 2000 1000 1000 5000 500 500 1000 500 500

0 0 0 0 0 0

N N N TN TN TN TN TN TN TN TN T TN Relativeabundance(% Relativeabundance(%) Relativeabundance(% Relativeabundance(%) Relativeabundance(%) -NTN H -H Relativeabundance(%) -NTN H -HT -NT -H -NTN H -HTN -NTN H -H -NTN -H D -p D D -p D D -pH D D -p D D -p D D -pH D S D S S D S S D S S D S S D S S D S K S K K S K K S K K S K K S K K S K K K K K K K Fig. 5 Comparison of functional pathways between three stages of blood pressure in KSD patients. Gene functions were predicted from 16S rRNA gene-based microbial compositions using the PICRUSt algorithm and the Kyoto Encyclopedia of Genes and Genomes database. Wilcoxon rank-sum test was used to compare the diference of abundance between two groups. *, ** means p < 0.05, p < 0.01, respectively. HTN hypertension, KSD kidney stone disease, NTN normotension, pHTN pre-hypertension

with bladder cancer had increased bacterial richness the blood pressure have a direct consequence on the uri- compared to HCs [29]. nary microbiome. Terefore, early management of HTN Our PCoA results demonstrated that the microbi- in KSD patients might restore the urinary microbiome ome composition difered between HCs’ bladder urine homeostasis. (obtained by transurethral catheterization) and KSD Similar to the PCoA results and Venn diagram, the patients’ kidney pelvis urine. In addition, regarding the comparisons of bacterial phyla and genera also revealed three KSD groups, the microbiome composition difered that the HCs had a unique bacterial profle compared to only between the KSD-NTN and KSD-pHTN or KSD- the KSD patients. For example, Gardnerella (a genera HTN groups, suggesting that urinary dysbiosis begins in in the major phylum Actinobacteria in the HC group) the pHTN stage. Similar to the PCoA results, the num- accounted for nearly 20% of the total bacterial genera in bers of OTUs shared by the HC group and the three KSD the HC group, but it only represented a very small pro- groups were lower than the numbers shared by the KSD- portion in the three KSD groups. Tis might be due to NTN and KSD-pHTN groups, the KSD-NTN and KSD- the fact that the HCs’ bladder urine samples (obtained HTN groups and the KSD-pHTN and KSD-HTN groups. by transurethral catheterization) may have been slightly In addition, the number of OTUs shared by the KSD- contaminated by bacteria living in the , as a pre- NTN and KSD-pHTN groups was slightly lower than the vious study demonstrated that Gardnerella was slightly number shared by the KSD-NTN and KSD-HTN groups higher in bladder urine samples obtained by catheteri- and by the KSD-pHTN and KSD-HTN groups. Moreo- zation than in urine samples obtained by suprapubic ver, some bacterial phyla and genera in the KSD groups aspiration [30]. In the present study, the ureteroscope were not detectable in the HC group and some bacterial for collecting kidney pelvis urine samples was inserted phyla and genera were detected in just one or two of the through the bladder but the bladder was repeatedly disin- KSD groups. Among the KSD groups, the highest or low- fected with iodophor and normal saline, so the risk of the est levels of Acidobacteria, Deinococcus–thermus, Fuso- pelvis urine being contaminated was lower than the risk bacteria, Blastococcus, Delftia and Lactobacillus were of the bladder urine being contaminated. observed in the KSD-pHTN group. Tese fndings indi- Interestingly, compared to the HC group, the KSD- cate that the urinary microbiome exhibits changes start- NTN group had a higher level of Proteobacteria and ing at the pHTN stage and it continues to evolve during higher levels of the associated genera Acinetobac- progression to HTN. Te parabolic trends of certain bac- ter, Comamonas and Delftia, which are considered teria (Bifdobacterium, Cloacibacterium, Lactobacillus, to be pathogens [31–33]. Te KSD-NTN group also etc.) across the three BP stages suggest that changes in had a higher level of Firmicutes and a higher level of Liu et al. J Transl Med (2020) 18:130 Page 10 of 13

the associated genus Lactobacillus. Lactobacillus can of Parasutterella in individuals with HTN [44] and in decrease the excretion of urinary [34], which individuals with NTN [45], it cannot be concluded that is responsible for the formation of KSD [35]. Tus, the the observed change in Parasutterella abundance in the increase in probiotic bacteria such as Lactobacillus urine is diferent from that in the gut, as both previous accompanying the increase in pathogenic bacteria in the studies only separated the participants into NTN and KSD-NTN group (compared to the HC group) might be HTN cohorts, and it is possible that the former included a self-protective response in the urinary microbiome. individuals with pHTN [44, 45]. Stenotrophomonas was Similar fndings were reported by Siddiqui et al. and enriched in the KSD-NTN group relative to the other in our previous study [36, 37]. In the study by Siddiqui two KSD groups in our study, but it is not reasonable to et al., more than 90% of the sequence reads in the urine conclude that it is not associated with the pathogenesis of of patients with belonged to the genus HTN as other studies have reported that it is associated Lactobacillus, a marked increase compared to 60% in the with poor health status. For example, in women undergo- urine of HCs [36]. Additionally, we previously found that ing pelvic foor surgery, Stenotrophomonas was detected patients with type 2 had a higher level of Lacto- at a higher rate in participants with positive urine cul- bacillus compared to HCs [37]. ture on the day of surgery and/or post-operatively than Among the three KSD groups, the KSD-HTN group in participants with a negative culture [46]. Additionally, had the lowest level of Proteobacteria and the associated in a small case–control study, women with urgent uri- genus Acidovorax, which has been shown to have high nary incontinence had an increased abundance of Steno- abundance in urine from patients with UTI and infam- trophomonas [47]. mation [38]. In contrast, the KSD-HTN group had the Interestingly, several bacterial genera in the phylum highest levels of Actinobacteria and the associated genus Firmicutes showed similar patterns to those previously Bifdobacterium and the highest levels of Firmicutes and reported in the gut microbiome of HTN patients. For the associated genus Lactobacillus. In addition, Lacto- instance, Blautia was identifed as a non-HTN-associ- bacillus abundance was positively correlated with BP. ated bacterial genus in the present study (it exhibited Infammation contributes to elevated BP [39–41], and decreased abundance in the urine of the KSD-NTN Bifdobacterium and Lactobacillus infuence the host group) and Blautia hansenii, was previously reported to infammatory response [42]. Terefore, the elevated lev- be a non-HTN-associated bacteria in the gut microbiome els of Actinobacteria and the associated genus Bifdo- [48]. Additionally, Butyrivibrio, which has been shown bacterium and the elevated levels of Firmicutes and the to be underrepresented in the gut of HTN patients [16], associated genus Lactobacillus might be involved in a declined in the KSD-HTN comparing to the HCs. self-regulatory response in KSD patients with co-occur- Tere were fewer genera with signifcant diferences ring HTN. in abundance between the KSD-pHTN and KSD-HTN Te Fusobacteria level was markedly higher in the groups than between the KSD-NTN and KSD-pHTN KSD-pHTN group than in the KSD-NTN group and groups or the KSD-NTN and KSD-HTN groups. Most slightly higher in the KSD-HTN group than in the KSD- of the genera belonging to Proteobacteria, such as Aurei- NTN group. It was previously shown that Fusobacteria monas and Cardiobacterium, were decreased in the was enriched in the gut of pHTN and HTN patients [16], KSD-HTN group compared to the KSD-pHTN group. suggesting that these bacteria contribute to increased BP. Although there are rarely reports on their roles in human For the frst time, we identifed several genera belong- microbiome, Aureimonas spp. was found to be associated ing to the phylum Proteobacteria in human urine that with peritonitis [49] and Cardiobacterium spp. was found difered in relative abundance between the KSD-NTN to be associated with endocarditis [50]. and KSD-pHTN groups. For example, both the gen- Interestingly, alterations in metabolic pathways were era Ochrobactrum and Rhizobium were enriched in the also observed. For example, the KSD-pHTN and KSD- KSD-NTN group compared to the KSD-pHTN and KSD- HTN groups exhibited decreased nitrogen metabolism HTN groups. In a previous study, Ochrobactrum spp. was relative to the KSD-NTN group. Additionally, in a study isolated by EQUC in an analysis of the urine of calcium of the gut microbiome, three single-nucleotide polymor- stone patients and Rhizobium was detected in a stone in phisms in the carbamoyl synthetase 1 gene one patient [43]. Tese fndings suggest that Ochrobac- (encoding a component of the nitrogen metabolism path- trum and Rhizobium are associated with KSD. Another way) were positively associated with persistent pulmo- genus in the phylum Proteobacteria, Parasutterella, was nary HTN in newborns [51]. However, there are likely higher in the KSD-pHTN group than in the KSD-NTN to be diferent mechanisms underlying functional difer- and KSD-HTN groups. Although previous studies of the ences between the urinary and gut microbiomes. Nucleo- human gut microbiome reported an increased abundance tide metabolism was also enriched in the KSD-pHTN and Liu et al. J Transl Med (2020) 18:130 Page 11 of 13

KSD-HTN groups compared to the KSD-NTN group. Additional fle 1: Figure S1. Comparison of the distribution of bacterial Previous studies demonstrated that nucleotide phylum between HC and KSD-NTN, between HC and KSD-pHTN, between metabolism disorder can lead to abnormal serum uric HC and KSD-HTN, between KSD-NTN and KSD-pHTN, between KSD-pHTN and KSD-HTN, using Wilcox rank-sum test. Horizontal bar represents acid levels, which contribute to the onset and progression mean and error bar represents SD. Bacterial phyla showing signifcantly of both KSD and HTN [52–54]. Tus, the increased activ- diferent abundance between the± two groups are shown. *p < 0.05; **p < ity of the nucleotide metabolism pathway associated with 0.01. Abbreviations: KSD: kidney stone disease; HC: healthy controls; HTN: hypertension; NTN: normotension; pHTN: pre-hypertension. the urinary microbiome may underlie the co-occurrence Additional fle 2: Table S1. Comparison of bacterial genus abundance of KSD and HTN. Finally, the adherens junction pathway in groups. Wilcoxon rank-sum test was used to compare the diference of was enriched in the KSD-NTN group compared to the abundance between two groups, and a, b, c means that there was signif- KSD-HTN group. In a previous study in rats, expression cant diference between groups of HC and KSD-NTN, between groups HC and KSD-pHTN, between groups of HC and KSD-HTN (p < 0.05). Abbrevia- of the adherens junction E-cadherin was shown tions: HC, healthy controls; HTN, hypertension; KSD, kidney stone disease; to be negatively correlated with BP [55]. Tese results NTN, normotension; pHTN, pre-hypertension. indicate that the adherens junction pathways of the uri- Additional fle 3: Table S2. Comparison of functional pathways in nary microbiome may be involved in the occurrence or groups. Gene functions were predicted from 16S rRNA gene-based microbial compositions using the PICRUSt algorithm and the Kyoto Ency- development of KSD-HTN. clopedia of Genes and Genomes database. Wilcoxon rank-sum test was Our study had several limitations. First, the sample used to compare the diference of abundance between two groups, and size was small and the study was not powered to detect a, b, c, d, e,f means that there was signifcant diference between groups of HC and KSD-NTN, between groups HC and KSD-pHTN, between groups of a signifcant association between HTN and the urinary HC and KSD-HTN, between KSD-NTN and KSD-pHTN, between KSD-NTN microbiome profle. Second, we only collected bladder and KSD-HTN, between KSD-pHTN and KSD-HTN (p < 0.05). Abbreviations: urine (obtained by transurethral catheterization) in the HC, healthy controls; HTN, hypertension; KSD, kidney stone disease; NTN, HC group to compare the urinary microbiome between normotension; pHTN, pre-hypertension. HCs and KSD patients. Tis was due to the invasiveness of the collection method used in the KSD patients (which Abbreviations KSD: Kidney stone disease; BP: Blood pressure; CT: Computed tomography; was conducted during the ureteroscopic lithotripsy pro- DBP: Diastolic blood pressure; HC: Healthy controls; HTN: Hypertension; KSD- cedure that was performed to break down the kidney HTN: KSD co-occurring with HTN; NTN: Normotension; pHTN: Prehypertension; stones) and resultant ethical considerations. Te use of SBP: Systolic blood pressure; UTI: Urinary tract infection. the diferent sample collection method in HCs meant that Acknowledgements we could not confrm that the diferences observed were We gratefully acknowledge the volunteers who participated in our study and not due to diferences between bladder urine and kid- the surgeons who collected the kidney pelvis urine samples. ney pelvis urine. Tird, as the number of males and body Authors’ contributions mass index difered signifcantly among the four groups, Conceptualization, N-HF, X-YX, L-XL and F-PL; methodology, NZ, PJ, N-HF, CL, and because of the high prevalence of KSD in males and D-SY, YW, Y-WZ and X-YX; software, F-PL and Q-XZ; validation, F-PL and N-HF; formal analysis, F-PL; investigation, NZ and PJ; resources, N-HF and X-YX; data overweight individuals, we cannot rule out confounding curation, F-PL, NZ and Y-WZ; writing—original draft preparation, F-PL; writ- related to these factors [56, 57]; stratifed sampling with a ing—review and editing, F-PL and Q-XZ; visualization, F-PL; supervision, F-PL larger sample size is needed in future research. and N-HF; project administration, N-HF and F-PL; funding acquisition, N-HF. All authors read and approved the fnal manuscript.

Conclusions Funding In summary, the results of this study demonstrate that This research was funded by the Special Project of Medical Innovation Team in Jiangsu Province (Grant number: CXTDA2017047). The funding body had no changes in the urinary microbiome profle occur in KSD role in the design of the study, the collection, analysis, or interpretation of data patients progressing from NTN to HTN and that there or in writing the manuscript. are metabolic pathway alterations in patients with abnor- Availability of data and materials mal BP. Tese fndings provide insight into the potential Sequencing data from this study have been deposited in the GenBank use of the urinary microbiome as a tool for monitoring Sequence Read Archive under Accession number PRJNA561017 (https​://www. and managing KSD co-occurring with HTN. For exam- ncbi.nlm.nih.gov/biopr​oject​/PRJNA​56101​7/). ple, lifestyle changes known to infuence the microbiome Ethics approval and consent to participate could be initiated at the pHTN stage to prevent progres- The study was approved by the Wuxi Second People’s Hospital, Nanjing sion to HTN. Medical University (ref. 201802) and informed consent was obtained from all subjects.

Supplementary information Consent for publication Supplementary information accompanies this paper at https​://doi. Not applicable. org/10.1186/s1296​7-020-02282​-3. Competing interests The authors declare that they have no competing interests. Liu et al. J Transl Med (2020) 18:130 Page 12 of 13

Author details 16. Li J, Zhao F, Wang Y, Chen J, Tao J, Tian G, Wu S, Liu W, Cui Q, Geng B, et al. 1 Wuxi School of Medicine, Jiangnan University, Wuxi, Jiangsu, China. 2 Depart- Gut microbiota dysbiosis contributes to the development of hyperten- ment of Urology, Afliated Wuxi No. 2 Hospital, Nanjing Medical University, sion. Microbiome. 2017;5:14. Wuxi, Jiangsu, China. 3 State Key Laboratory of Food Science and Technol- 17. Yang T, Santisteban MM, Rodriguez V, Li E, Ahmari N, Carvajal JM, Zadeh ogy and School of Food Science and Technology, Jiangnan University, Wuxi, M, Gong M, Qi Y, Zubcevic J, et al. Gut dysbiosis is linked to hypertension. Jiangsu, China. 4 State Key Laboratory for Diagnosis and Treatment of Infec- Hypertension. 2015;65:1331–40. tious Diseases, National Clinical Research Center for Infectious Diseases, 18. Ko CY, Hu AK, Chou D, Huang LM, Su HZ, Yan FR, Zhang XB, Zhang HP, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Zeng YM. Analysis of oral microbiota in patients with obstructive sleep Diseases, The First Afliated Hospital, College of Medicine, Zhejiang University, apnea-associated hypertension. Hypertens Res. 2019;42:1692–700. Hangzhou, Zhejiang, China. 19. Blacher E, Levy M, Tatirovsky E, Elinav E. Microbiome-modulated metabo- lites at the interface of host immunity. J Immunol. 2017;198:572–80. Received: 19 October 2019 Accepted: 26 February 2020 20. Booth JN, Li J, Zhang L, Chen L, Muntner P, Egan B. Trends in prehy- pertension and hypertension risk factors in US adults. Hypertension. 2017;70:275–84. 21. Thomas-White K, Forster SC, Kumar N, Van Kuiken M, Putonti C, Stares MD, Hilt EE, Price TK, Wolfe AJ, Lawley TD. Culturing of female bladder References bacteria reveals an interconnected urogenital microbiota. Nat Commun. 1. Fink HA, Wilt TJ, Eidman KE, Garimella PS, MacDonald R, Rutks IR, 2018;9:1557. Brasure M, Kane RL, Ouellette J, Monga M. Medical management to 22. Ling Z, Liu F, Shao L, Cheng Y, Li L. Dysbiosis of the urinary microbiota prevent recurrent nephrolithiasis in adults: a systematic review for an associated with urine levels of proinfammatory chemokine interleukin-8 American College of Physicians Clinical Guideline. Ann Intern Med. in female type 2 diabetic patients. Front Immunol. 2017;8:1032. 2013;158:535–43. 23. Magoc T, Salzberg SL. FLASH: fast length adjustment of short reads to 2. Kittanamongkolchai W, Mara KC, Mehta RA, Vaughan LE, Denic A, improve genome assemblies. Bioinformatics. 2011;27:2957–63. Knoedler JJ, Enders FT, Lieske JC, Rule AD. Risk of hypertension among 24. Davis NM, Proctor DM, Holmes SP, Relman DA, Callahan BJ. Simple statisti- frst-time symptomatic kidney stone formers. Clin J Am Soc Nephrol. cal identifcation and removal of contaminant sequences in marker-gene 2017;12:476–82. and metagenomics data. Microbiome. 2018;6:226. 3. El-Zoghby ZM, Lieske JC, Foley RN, Bergstralh EJ, Li X, Melton LR, Kram- 25. Langille MG, Zaneveld J, Caporaso JG, McDonald D, Knights D, Reyes JA, beck AE, Rule AD. Urolithiasis and the risk of ESRD. Clin J Am Soc Nephrol. Clemente JC, Burkepile DE, Vega TR, Knight R, et al. Predictive func- 2012;7:1409–15. tional profling of microbial communities using 16S rRNA marker gene 4. Ku E, Lee BJ, Wei J, Weir MR. Hypertension in CKD: core curriculum 2019. sequences. Nat Biotechnol. 2013;31:814–21. Am J Kidney Dis. 2019;74:120–31. 26. Min Y, Ma X, Sankaran K, Ru Y, Chen L, Baiocchi M, Zhu S. Sex-specifc 5. Eisner BH, Porten SP, Bechis SK, Stoller ML. Hypertension is associated association between gut microbiome and fat distribution. Nat Commun. with increased urinary calcium excretion in patients with nephrolithiasis. 2019;10:1–9. J Urol. 2010;183:576–9. 27. Wu P, Chen Y, Zhao J, Zhang G, Chen J, Wang J, Zhang H. Urinary microbi- 6. Villa-Etchegoyen C, Lombarte M, Matamoros N, Belizán JM, Cormick G. ome and psychological factors in women with overactive bladder. Front Mechanisms involved in the relationship between low calcium intake Cell Infect Microbiol. 2017;7:488. and high blood pressure. Nutrients. 2019;11:1112. 28. Thomas-White KJ, Hilt EE, Fok C, Pearce MM, Mueller ER, Kliethermes S, 7. Karbach SH, Schonfelder T, Brandao I, Wilms E, Hormann N, Jackel S, Jacobs K, Zilliox MJ, Brincat C, Price TK, et al. Incontinence medication Schuler R, Finger S, Knorr M, Lagrange J, et al. Gut microbiota promote response relates to the female urinary microbiota. Int Urogynecol J. angiotensin II-induced arterial hypertension and vascular dysfunction. J 2016;27:723–33. Am Heart Assoc. 2016;5:e3698. 29. Wu P, Zhang G, Zhao J, Chen J, Chen Y, Huang W, Zhong J, Zeng J. Profl- 8. Cheema MU, Pluznick JL. Gut microbiota plays a central role to modulate ing the urinary microbiota in male patients with bladder cancer in China. the plasma and fecal metabolomes in response to angiotensin II. Hyper- Front Cell Infect Microbiol. 2018;8:167. tension. 2019;74:184–93. 30. Wolfe AJ, Toh E, Shibata N, Rong R, Kenton K, FitzGerald M, Mueller 9. Karstens L, Asquith M, Caruso V, Rosenbaum JT, Fair DA, Braun J, Gregory ER, Schreckenberger P, Dong Q, Nelson DE, Brubaker L. Evidence of WT, Nardos R, McWeeney SK. Community profling of the urinary uncultivated bacteria in the adult female bladder. J Clin Microbiol. microbiota: considerations for low-biomass samples. Nat Rev Urol. 2012;50:1376–83. 2018;15:735–49. 31. Wong D, Nielsen TB, Bonomo RA, Pantapalangkoor P, Luna B, Spellberg B. 10. Shrestha E, White JR, Yu SH, Kulac I, Ertunc O, De Marzo AM, Yegnasu- Clinical and pathophysiological overview of Acinetobacter infections: a bramanian S, Mangold LA, Partin AW, Sfanos KS. Profling the urinary century of challenges. Clin Microbiol Rev. 2017;30:409–47. microbiome in men with positive versus negative biopsies for prostate 32. Almuzara MN, Cittadini R, Ocampo CV, Bakai R, Traglia G, Ramirez MS, Cas- cancer. J Urol. 2018;199:161–71. tillo MD, Vay CA. Intra-Abdominal infections due to Comamonas kerstersii. 11. Gottschick C, Deng Z, Vital M, Masur C, Abels C, Pieper DH, Wagner- J Clin Microbiol. 2013;51:1998–2000. Döbler I. The urinary microbiota of men and women and its changes in 33. Bilgin H, Sarmis A, Tigen E, Soyletir G, Mulazimoglu L. Delftia acidovorans: women during bacterial vaginosis and antibiotic treatment. Microbiome. a rare pathogen in immunocompetent and immunocompromised 2017;5:99. patients. Can J Infect Dis Med Microbiol. 2015;26:277–9. 12. Fouts DE, Pieper R, Szpakowski S, Pohl H, Knoblach S, Suh MJ, Huang 34. Ferraz RR, Marques NC, Froeder L, Menon VB, Siliano PR, Baxmann ST, Ljungberg I, Sprague BM, Lucas SK, et al. Integrated next-generation AC, Heilberg IP. Efects of Lactobacillus casei and Bifdobacterium breve sequencing of 16S rDNA and metaproteomics diferentiate the healthy on urinary oxalate excretion in nephrolithiasis patients. Urol Res. urine microbiome from asymptomatic bacteriuria in neuropathic bladder 2009;37:95–100. associated with spinal cord . J Transl Med. 2012;10:174. 35. Mitchell T, Kumar P, Reddy T, Wood K, Knight J, Assimos D, Holmes R. 13. Tisler A, Pierratos A, Honey JD, Bull SB, Rosivall L, Logan AG. High urinary Dietary oxalate and kidney stone formation. Am J Physiol Renal Physiol. excretion of uric acid combined with high excretion of calcium links 2019;316:409–13. kidney stone disease to familial hypertension. Nephrol Dial Transplant. 36. Siddiqui H, Lagesen K, Nederbragt AJ, Jeansson SL, Jakobsen KS. Altera- 2002;17:253–9. tions of microbiota in urine from women with interstitial cystitis. BMC 14. Pfau A, Knauf F. Update on nephrolithiasis: core curriculum 2016. Am J Microbiol. 2012;12:205. Kidney Dis. 2016;68:973–85. 37. Liu F, Ling Z, Xiao Y, Lv L, Yang Q, Wang B, Lu H, Zheng L, Jiang P, Wang W, 15. DiBianco JM, Jarrett TW, Mufarrij P. and nephrolithi- Li L. Dysbiosis of urinary microbiota is positively correlated with type 2 asis risk: should the medical management of nephrolithiasis include the diabetes mellitus. Oncotarget. 2017;8:3798–810. treatment of metabolic syndrome? Rev Urol. 2015;17:117–28. Liu et al. J Transl Med (2020) 18:130 Page 13 of 13

38. Moustafa A, Li W, Singh H, Moncera KJ, Torralba MG, Yu Y, Manuel O, Biggs urgency urinary incontinence and its severity? Front Cell Infect Microbiol. W, Venter JC, Nelson KE, et al. Microbial metagenome of urinary tract 2016;6:78. infection. Sci Rep. 2018;8:1–12. 48. Yan Q, Gu Y, Li X, Yang W, Jia L, Chen C, Han X, Huang Y, Zhao L, Li P, et al. 39. Tomiyama H, Shiina K, Matsumoto-Nakano C, Ninomiya T, Komatsu S, Alterations of the gut microbiome in hypertension. Front Cell Infect Kimura K, Chikamori T, Yamashina A. The contribution of infammation Microbiol. 2017;7:381. to the development of hypertension mediated by increased arterial stif- 49. Schröttner P, Rudolph WW, Taube F, Gunzer F. First report on the isolation ness. J Am Heart Assoc. 2017;6:e005729. of Aureimonas altamirensis from a patient with peritonitis. Int J Infect Dis. 40. Bautista LE, Vera LM, Arenas IA, Gamarra G. Independent association 2014;29:71–3. between infammatory markers (C-reactive protein, interleukin-6, 50. Malani A, Aronof D, Bradley S, Kaufman C. Cardiobacterium hominis and TNF-alpha) and essential hypertension. J Hum Hypertens. endocarditis: two cases and a review of the literature. Eur J Clin Microbiol 2005;19:149–54. Infect Dis. 2006;25:587–95. 41. Sesso HD, Buring JE, Rifai N, Blake GJ, Gaziano JM, Ridker PM. 51. Kaluarachchi DC, Smith CJ, Klein JM, Murray JC, Dagle JM, Ryckman C-reactive protein and the risk of developing hypertension. JAMA. KK. Polymorphisms in urea cycle enzyme genes are associated with 2003;290:2945–51. persistent pulmonary hypertension of the newborn. Pediatr Res. 42. Carroll IM, Andrus JM, Bruno-Bárcena JM, Klaenhammer TR, Hassan HM, 2018;83:142–7. Threadgill DS. Anti-infammatory properties of Lactobacillus gasseri 52. Shankar A, Klein R, Klein BEK, Nieto FJ. The association between serum expressing manganese superoxide dismutase using the interleukin uric acid level and long-term of hypertension: population- 10-defcient mouse model of colitis. Am J Physiol Gastrointest based cohort study. J Hum Hypertens. 2006;20:937–45. Physiol. 2007;293:729–38. 53. Zhang W, Sun K, Yang Y, Zhang H, Hu FB, Hui R. Plasma uric acid and 43. Dornbier RA, Bajic P, Van Kuiken M, Jardaneh A, Lin H, Gao X, Knudsen B, hypertension in a Chinese community: prospective study and metaanaly- Dong Q, Wolfe AJ, Schwaderer AL. The microbiome of calcium-based uri- sis. Clin Chem. 2009;55:2026–34. nary stones. Urolithiasis. 2019. https​://doi.org/10.1007/s0024​0-019-01146​ 54. Coe FL, Kavalach AG. and in patients with -w. calcium nephrolithiasis. N Engl J Med. 1974;291:1344. 44. Mushtaq N, Hussain S, Zhang S, Yuan L, Li H, Ullah S, Wang Y, Xu J. 55. Wang Y, Mu J, Liu F, Ren K, Xiao H, Yang Z, Yuan Z. Salt-induced epithelial- Molecular characterization of alterations in the intestinal microbiota of to-mesenchymal transition in Dahl salt-sensitive rats is dependent on patients with grade 3 hypertension. Int J Mol Med. 2019;44:513–22. elevated blood pressure. Braz J Med Biol Res. 2014;47:223–30. 45. Dan X, Mushi Z, Baili W, Han L, Enqi W, Huanhu Z, Shuchun L. Diferential 56. Scales CD, Smith AC, Hanley JM, Saigal CS. Prevalence of kidney stones in analysis of hypertension-associated intestinal microbiota. Int J Med Sci. the . Eur Urol. 2012;62:160–5. 2019;16:872–81. 57. Del VE, Negri AL, Spivacow FR, Rosende G, Forrester M, Pinduli I. Meta- 46. Nienhouse V, Gao X, Dong Q, Nelson DE, Toh E, McKinley K, Schreck- bolic diagnosis in stone formers in relation to body mass index. Urol Res. enberger P, Shibata N, Fok CS, Mueller ER, et al. Interplay between 2012;40:47–52. bladder microbiota and urinary antimicrobial peptides: mechanisms for human urinary tract infection risk and symptom severity. PLoS ONE. 2014;9:e114185. Publisher’s Note 47. Karstens L, Asquith M, Davin S, Staufer P, Fair D, Gregory WT, Rosenbaum Springer Nature remains neutral with regard to jurisdictional claims in pub- JT, McWeeney SK, Nardos R. Does the urinary microbiome play a role in lished maps and institutional afliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions